Solvent-Based Solubility Predictions in Pharmaceutical Systems

Solubility prediction in pharmaceutical development underlies formulation design, bioavailability optimisation and process scale-up. Solvent-based approaches harness thermodynamic models and empirical correlations to estimate drug solubility in individual or mixed solvents, thereby reducing experimental burden. Early frameworks such as log-linear and Yalkowsky cosolvency models provide rapid estimates using aqueous solubility and cosolvent fraction, while semi-empirical methods (for example, the Jouyban–Acree model) incorporate additional data points to enhance accuracy. Contemporary efforts integrate molecular thermodynamics and machine-learning techniques to capture solute–solvent interactions at the molecular level, leveraging parameters such as solubility parameters, activity coefficients and excess thermodynamic functions. These predictive tools guide the selection of green solvents and co-formers, underpinning sustainable processing and enabling formulation of poorly soluble compounds. Advances in calorimetric measurement and equation-of-state modelling have expanded predictive scope across temperature and composition spaces, offering broad applicability to drug discovery and manufacturing on a global scale.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Recent advances include the use of fast scanning calorimetry combined with PC-SAFT to measure melting enthalpies of amino acids at ultrahigh heating rates, enabling accurate prediction of aqueous solubility across a broad temperature span. Computational studies on a poorly soluble pyridazinone derivative in DMSO–water binary systems have applied multiple cosolvency models (including van’t Hoff, Apelblat and Jouyban–Acree) to correlate solubility data and thermodynamic parameters with deviations under 2 % RMSD, elucidating enthalpy–entropy contributions to dissolution. Furthermore, inverse gas chromatography paired with Hansen solubility parameters has been employed to derive Hildebrand and Hansen metrics for ionic liquids, facilitating solvent screening for drug solubilisation by quantifying dispersion, polar and hydrogen-bonding interactions.

Solvent-Based Solubility Predictions in Pharmaceutical Systems publication trend

The graph below shows the total number of articles in solvent-based solubility predictions in pharmaceutical systems across all publications each year (not limited to Nature Index journals).

Technical terms

Cosolvency model: A mathematical framework describing solute solubility in mixtures of two or more solvents based on solvent composition and interaction parameters.

Hansen solubility parameters: A set of three values (dispersion, polar and hydrogen-bonding) that predict solute–solvent affinity and miscibility by partitioning cohesive energy density.

PC-SAFT: An advanced equation of state that models phase behaviour of associating fluids by accounting for molecular size, shape and specific interactions.

Fast scanning calorimetry (FSC): A technique that measures thermal transitions at exceptionally high heating rates to determine accurate melting properties of thermally labile compounds.

Activity coefficient: A dimensionless factor quantifying the deviation of a solute’s behaviour from ideality in a given solvent mixture, reflecting molecular interactions.

References

1. Review of the cosolvency models for predicting solubility of drugs in water-cosolvent mixtures.. Journal of Pharmacy & Pharmaceutical Sciences (2008).
2. New experimental melting properties as access for predicting amino-acid solubility. RSC Advances (2018).
3. Experimental and Computational Approaches for Solubility Measurement of Pyridazinone Derivative in Binary (DMSO + Water) Systems. Molecules (2019).
4. Practical Determination of the Solubility Parameters of 1-Alkyl-3-methylimidazolium Bromide ([CnC1im]Br, n = 5, 6, 7, 8) Ionic Liquids by Inverse Gas Chromatography and the Hansen Solubility Parameter. Molecules (2019).